google / lightweight_mmm

LightweightMMM 🦇 is a lightweight Bayesian Marketing Mix Modeling (MMM) library that allows users to easily train MMMs and obtain channel attribution information.
https://lightweight-mmm.readthedocs.io/en/latest/index.html
Apache License 2.0
858 stars 178 forks source link

Can we have different number of records for different geos ? #201

Open shekharkhandelwal1983 opened 1 year ago

shekharkhandelwal1983 commented 1 year ago

I have a scenario where I have to build a heirarchical Bayesian model with 46 geos, and the number of records for each geo are different. Is LWMMM modelling mechanism allows to model such datasets ? if yes, then how to transform the dataset for the modelling?

uomodellamansarda commented 1 year ago

For better understanding why you need 46 Geos? What is the business scenario?

shekharkhandelwal1983 commented 1 year ago

@uomodellamansarda I am building a Bayesian hierarchical modelling, and the dataset contains records for 46 different regions. Its a normal MMM scenario, just for multiple regions. Happy to share more details, if required.